Facile conversion of the surface layers of graphite to capacitive manganese oxide coatings

Article abstract of DOI:10.1149/1.2868767

A simple method to convert graphite to manganese oxide coatings is presented in this paper. The as-grown coatings were characterized by X-ray photoelectron spectroscopy, X-ray diffraction, scanning electron microscopy, and energy-dispersive X-ray analyses, and thus verified to be largely amorphous manganese oxide. The coatings behave typically capacitive in various neutral aqueous solutions of alkali metal salts, e.g., 0.5 M LiCl, 0.5 M NaCl, and 0.5 M KCl. The capacitance per geometric electrode surface area (Ca) from cyclic voltammetry in 0.5 M NaCl increased to 67.2 mF/ cm2 when the conversion time was up to 120 min, obeying the logarithm law very well. In addition, the capacitance per mass of the coating (Cm) of 362, 385, 380, and 374 F g-1 was achieved with the manganese coatings grown for 10, 30, 60, and 120 min, respectively. Electrochemical measurements demonstrated evident differences in capacitive behaviors of the coatings in the three above-mentioned electrolyte solutions. It was confirmed that this facile conversion of the surface layers of graphite offers a simple and efficient formation of thin-film supercapacitors promising for industrial applications.

Full text of DOI:10.1149/1.2868767

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Journal of The Electrochemical Society, 155 ͑5͒ A355-A360 ͑2008͒
A355
0013-4651/2008/155͑5͒/A355/6/$23.00 © The Electrochemical Society
Facile Conversion of the Surface Layers of Graphite to
Capacitive Manganese Oxide Coatings
Mengqiang Wu,a,z Liping Zhang,b Jiahui Gao,a Ying Zhou,a Shuren Zhang,a and
Ai Chena
aState Key Laboratory of Electronic Thin Films and Integrated Devices, University of Electronic Science
and Technology of China, Chengdu 610054, China
bDepartment of Materials and Chemical Engineering, Sichuan University of Science and Engineering,
Zigong 643000, China
A simple method to convert graphite to manganese oxide coatings is presented in this paper. The as-grown coatings were
characterized by X-ray photoelectron spectroscopy, X-ray diffraction, scanning electron microscopy, and energy-dispersive X-ray
analyses, and thus verified to be largely amorphous manganese oxide. The coatings behave typically capacitive in various neutral
aqueous solutions of alkali metal salts, e.g., 0.5 M LiCl, 0.5 M NaCl, and 0.5 M KCl. The capacitance per geometric electrode
surface area ͑Ca͒ from cyclic voltammetry in 0.5 M NaCl increased to 67.2 mF/cm2 when the conversion time was up to 120 min,
obeying the logarithm law very well. In addition, the capacitance per mass of the coating ͑Cm͒ of 362, 385, 380, and 374 F g−1
was achieved with the manganese coatings grown for 10, 30, 60, and 120 min, respectively. Electrochemical measurements
demonstrated evident differences in capacitive behaviors of the coatings in the three above-mentioned electrolyte solutions. It was
confirmed that this facile conversion of the surface layers of graphite offers a simple and efficient formation of thin-film super-
capacitors promising for industrial applications.
© 2008 The Electrochemical Society. ͓DOI: 10.1149/1.2868767͔ All rights reserved.
Manuscript submitted May 21, 2007; revised manuscript received January 4, 2008. Available electronically March 11, 2008.
In recent decades, much attention has been focused on electro-
chemical supercapacitors based on pseudocapacitance from faradaic
redox reactions. Accordingly, there have been increasing attempts to
develop cost-effective pseudocapacitive materials instead of the pre-
cious metal ruthenium oxides,1,2 such as nonprecious metal oxides3-9
and nitrides,10-12 electroactive polymers,13,14 and their
bath at 50°C, and the capacitive performances of the as-grown coat-
ings were investigated comprehensively. The aim of this work is to
confirm that this facile conversion of the surface layers of graphite
facilitates the simple and efficient formation of promising thin-film
supercapacitors economical for industrial applications.
composites.15-17 Amorphous
hydrous
manganese
oxides
Experimental
͑a-MnO2·nH2O͒ recently have been attracting considerable interest
as the most promising electrode materials. Most of the efforts are
focused on depositing a-MnO2·nH2O thin films onto the electron
collectors. In a universally used sol–gel technique, the formation of
a hydrous MnO2 coating is achieved by a process involving the
Graphite disk electrodes were prepared in a procedure described
in detail elsewhere.31 The as-fabricated graphite electrodes were
placed vertically in a bath containing freshly prepared 0.25 M
KMnO4 in a solution of 0.5 M H2SO4, which was continuously
stirred during the conversion. The conversions were carried out at
50°C in the identical but fresh solutions for the desired time dura-
tions. Upon conversions, the coatings were well rinsed with distilled
water, cleaned ultrasonically for 30 min, and dried by a flow of cool
air. The scanning electron microscopic ͑SEM͒ images were recorded
with a JSM-5800LV microscope ͑JEOL, Japan͒, and the energy-
dispersive X-ray analyses ͑EDX͒ data were collected simulta-
neously. The X-ray diffraction ͑XRD͒ measurements were per-
formed on an X-ray diffractometer ͑X’PertProMPD, Philips Co.͒
using Cu K␣ radiation, and X-ray photoelectron spectroscopy ͑XPS͒
analysis was carried out on an XSAM 800 spectrometer equipped
with a Mg K␣ radiation. The mass of the grown manganese oxide
films was determined by flame atomic absorption spectrometry.
First, a manganese oxide film was thoroughly converted to a man-
ganese ionic solution of sample by treatment with 1.0 mol L−1 ni-
tric acid. An AA 6800 atomic absorption spectrometer ͑Shimadzu,
Japan͒ was utilized for the manganese measurements. A manganese
hollow cathode lamp operating at 6 mA was used as the radiation
source. The primary resonance line at 279.5 nm was applied with a
bandwidth of 0.5 nm. The acetylene flow rate and the burner height
were adjusted for a maximum absorbance signal by aspirating a
solution of the sample.
The coatings were comprehensively evaluated via cyclic voltam-
metry ͑CV͒, chronopotentiometric ͑CP͒, and electrochemical imped-
ance spectroscopy ͑EIS͒ with a CHI760B computer-controlled elec-
trochemical workstation. An alternate current ͑ac͒ potential
amplitude of 10 mV was employed for the impedance measurements
in the frequency range between 100 kHz and 0.1 Hz. All electro-
chemical measurements were conducted at room temperature in a
beaker-type electrochemical cell equipped with a coating-covered
graphite working electrode, a Pt wire counter electrode, and a Ag/
AgCl ͑3 M KCl͒ reference electrode. Various neutral aqueous elec-
trolyte solutions of alkali metal salts were utilized in the electro-
reduction of the permanganate anion ͑MnO−4͒ with a mild reducing
18-20
agent such as Mn2+
.
In addition, an anodic electrochemical deposition of
a-MnO2·nH2O was extensively reported, but the obtained materials
had not yet achieved a particularly high specific capacitance.21-26
More encouragingly, the potentiodynamically deposited amorphous
nanostructured form of hydrous MnO2 was found to achieve a spe-
cific capacitance of 482 F/g.27 Successively, microporous, amor-
phous nickel–manganese oxides and cobalt–manganese oxides were
also deposited at high scan rates with specific capacitance as high
as 621 and 498 F/g,28 respectively. Despite this, a specific capaci-
tance of only 125 F/g was achieved by Chuang and Hu29 for
the anodically deposited hydrous manganese–cobalt oxides
͓͑Mn,Co͒Ox·nH2O͔. Undoubtedly, the electrochemical capacitance
is dependent on the film-depositing conditions. It was also reported
that an electrochemical oxidation of a Mn/MnO film in a 1 M
Na2SO4 solution can cause the formation of an electrochemical ca-
pacitance, and the porous surface layer is likely responsible for the
capacitive tendencies of the films30.
In the preliminary work,31 a one-step controllable redox deposi-
tion method was briefly described, which was developed from the
viewpoint of the surface modification of a graphite electrode with
oxidative potassium permanganate ͑KMnO4͒ in an acidic aqueous
solution at room temperature. Most recently, several researches32-37
employed this simple route to fabricate structured MnO2–carbon
composites mainly for electrochemical supercapacitors. In the
present work, the surface layers of graphite disk electrodes were
converted to manganese oxide coatings in an acidic KMnO4 solution
z E-mail: mwu@uestc.edu.cn
Downloaded on 2014-05-11 to IP 205.208.51.61 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
A356
Journal of The Electrochemical Society, 155 ͑5͒ A355-A360 ͑2008͒
Figure 2. The XRD pattern for the coating formed by conversion in 0.25 M
KMnO4 + 1.0 M H2SO4 at 50°C for 30 min.
corresponding to different oxygen-containing species such as Mn–
O–H at 530 eV,40 and Mn–O–H at 532 eV.41 According to the areas
of the XPS peaks, the atomic ratio of manganese to oxygen can be
estimated to be 1: 2.12, which is in agreement with that suggested
by the EDX analysis ͑not shown here͒. Besides the sharp diffraction
peak at 26° associated with the graphite, the peaks at around 38, 56,
and 68° show a certain degree of fine crystalline structure, which
should be attributed to peaks ͑211͒, ͑600͒, and ͑002͒ of ␣-MnO2,
respectively. Nevertheless, the XRD pattern generally has a lack of
clear peaks, and a wide broadening, indicating the amorphous nature
of the coating. Without further heat treatment, the as-formed coating
is thus presumably amorphous manganese oxide. Actually, to purify
and open carbon nanotubes in high yield, KMnO4 has been consid-
ered to be the best oxidant in an acidic solution.42 It was also dem-
onstrated that permanganate was transferred to manganese ͑IV͒ ox-
ide as a by-product, but washed off. However, the evolution of CO2
gas was not noticed during the period of electrode immersion in this
work. This is very likely attributed to the continuous stirring having
removed the gaseous products before it formed a visible bubble on
the electrode surface. Nevertheless, a vigorous bubbling did occur
when adding graphite powder into an identical conversion solution,
which is in accordance with Reaction 1
Figure 1. The XPS spectra for the coating formed by conversion in 0.25 M
KMnO4 + 1.0 M H2SO4 at 50°C for 30 min. ͑a͒ Spectrum at Mn 2p1/2 and
Mn 2p3/2, ͑b͒ spectrum at O 1s.
chemical characterizations, including 0.5 M LiCl, 0.5 M NaCl, and
0.5 M KCl. The specific capacitance was estimated on the basis of
the following equation: C = i/͑dV/dt͒, where i is the current re-
sponse of the corresponding CV plot, and dV/dt is the potential
sweep rate.
4KMnO4 + 3C + 2H2SO4→4MnO2 + 3CO2 + 2K2SO4 + H2O
͓1͔
Actually, Kim et al.32 clarified this reaction process in the system of
acetylene black and acidic KMnO4 solution via an in situ monitoring
of the reduction potential and pH of the solution and supplementary
ultraviolet-visible analysis. Because there likely exists a less com-
plete oxidation of carbon, as was demonstrated by Zoval et al.43 for
the formation of platinum particles on highly oriented pyrolytic
graphite, no attempt was made to estimate how much grew onto the
surface of the graphite by weighing the electrode before and after
the conversion process. But, basically, the mass of a coating can be
controlled by employing a different conversion time, and should
increase with the conversion time.
Results and Discussion
Micrographs of the as-formed coatings after 30 min conversion
were collected ͑not shown here͒. As verified in the previous work,31
the coating is densely and uniformly grown onto the surface of the
graphite electrode. No significant morphology difference was found
between the coatings grown at room temperature and at 50°C. How-
ever, growth at high temperature was accelerated observably. To
characterize the as-formed coatings, XPS and XRD analyses were
conducted and the results are shown in Fig. 1 and 2, respectively. As
shown in Fig. 1a, the Mn 2p3/2 peak is centered at 642.5 eV and the
Mn 2p1/2 peak at 654.2 eV, with a spin-energy separation of 11.7 eV,
which is in accordance with that reported in MnO2.38 It has also
been reported that the binding energies of the Mn 2p3/2 electron for
Mn3+ and Mn4+ are 641.6 and 642.6 eV,39 respectively. In addition,
no clear Mn 2p3/2 signal of KMnO4 at 647 eV can be observed,
confirming that the permanganate ions have been reduced to man-
ganese dioxide by the carbon in graphite. As shown in Fig. 1b, the O
1s spectra of the coatings could be fitted into two main constituents
Typical CVs measured in a 0.5 M NaCl aqueous solution for the
coatings formed by different conversion time are shown in Fig. 3.
These CVs clearly exhibit the rectangular and symmetric current-
potential characteristics of a capacitor, except at potentials near the
two limits of the potential window. Additionally, even at a high
potential scan rate of 1000 mV/s, the CVs were found to still show
a capacitive-like behavior, i.e., a rectangular-like shape. More sig-
nificantly, the voltammetric current response and thereby the capaci-
tance of the coating increase substantially with the conversion time.
As shown in Fig. 4, the capacitance per geometric electrode surface
Downloaded on 2014-05-11 to IP 205.208.51.61 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Journal of The Electrochemical Society, 155 ͑5͒ A355-A360 ͑2008͒
A357
Figure 5. The comparison of the voltammetric current of the coatings
formed by 10 and 60 min in three neutral aqueous solutions.
Figure 3. Cyclic voltammograms, measured in 0.5 M NaCl at 10 mV/s, of
the coatings formed by different time durations.
in this specific range. No attempt was therefore made to pursue the
maximum. For the sake of comparison between our results and those
in the literature, the capacitance per mass of the coating ͑Cm͒ was
also calculated with the mass evaluated using atomic absorption
spectroscopy. The average specific capacitances of 362, 385, 380,
and 374 F g−1 in a 0.5 M NaCl aqueous solution were achieved
from CV with the manganese coatings grown for 10, 30, 60, and 120
min, respectively. The results seem a little less than that with the
potentiodynamically deposited nanostructured manganese dioxide
films by Miura et al.,27,44 specifically, a maximum capacitance of
410−482 F g−1 at a scan rate of 10 mV s−1. It is, however, higher
than that with the anodically deposited hydrous manganese oxide
area ͑Ca͒, derived from the CV currents, increases to 67.2 mF/cm2
with the conversion time, up to 120 min, following the logarithm
law very well. Undoubtedly, the increase in capacitance is due to the
increase in Mn sites of the MnO2 coatings. Presumably, the loga-
rithm law could be ascribed to the diffusion process involved in the
conversion reactions. The data presented in Fig. 4 are averaged val-
ues over four scan rates, i.e., 5, 10, 25, and 50 mV/s. The capaci-
tance is significantly larger than that previously reported for the
coating grown at room temperature for the same conversion time.31
The logarithm law is as well expected to imply the dependence of
deposit mass or deposit thickness on the conversion time. As ex-
pected, no sharp interface was observed between the coating and the
graphite substrate, and the deposit thickness is not easily determined
at the observation of cross-section SEM imaging and EDX analysis.
Actually, these observations also confirm the origin of the MnO2
coating from the graphite substrate itself. Although admittedly it is
quite understandable, an attempt to make careful EDX analyses
should be encouraged and helpful to probe the interface between the
coating and the graphite substrate. Although the maximum specific
capacitance for the utmost long conversion time has not been
reached in this work, further improvements could be anticipated by
the above-mentioned logarithm law. The linear relationship between
the reaction extent and the logarithm of conversion time does hold
films by Hu et al.,21 i.e., 265−320 F g−1
.
The applicability of different neutral electrolytes for manganese
oxide-based electrode materials in supercapacitors has been
investigated.19,20,45 It was found that amorphous or crystallite man-
ganese oxide could be effectively used in several neutral electro-
lytes, and it was thus helpful for practical application and storage
safety. Figure 5 compares the voltammetric current responses of the
coatings formed by 10 and 60 min in three neutral aqueous solu-
tions. In order not to cause a significant difference in the electrical
resistance and alkali metal ionic concentration of the electrolyte, 0.5
M LiCl, 0.5 M NaCl, and 0.5 M KCl were employed in the electro-
chemical measurements. As shown in Fig. 5, a perceptible voltam-
metric current difference among the three different electrolytes is
well demonstrated for coatings formed by both 10 and 60 min, in-
dicating a clear similar trend for the two coatings that the voltam-
metric current is increasing with decreasing the size of the cation.
To further confirm the cation influence on the capacitive proper-
ties of the grown manganese oxide coatings, Fig. 6 compares the
charge–discharge plots of the coating formed by 10 min in different
electrolyte solutions at the same current density of 0.885 mA/cm2,
experimentally a current of 0.25 mA against the fixed graphite elec-
trode area. The anodic charging segments, in general, are symmetric
to their corresponding cathodic discharging counterparts for all three
plots, indicating a high reversibility of a typical capacitive material
in the three aqueous electrolyte solutions within the applied poten-
tial window, in accordance with the CV observations. The average
specific capacitance of an electrode in a given electrolyte solution
thus can be calculated via an equation as follows
i⌬t
C =
͓2͔
s⌬V
where i is the discharge current for the applied time duration ⌬t, ⌬V
is the potential window, and s is the fixed graphite electrode area.
Clearly, the charge–discharge plots in Fig. 6 indicate the difference
Figure 4. The logarithm relationship between the specific capacitance ͑Ca͒
and the conversion time.
Downloaded on 2014-05-11 to IP 205.208.51.61 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
A358
Journal of The Electrochemical Society, 155 ͑5͒ A355-A360 ͑2008͒
Figure 6. The charge–discharge plots of the coating formed by 10 min in
different electrolyte solutions at a current density of 0.885 mA/cm2.
Figure 7. Specific capacitance of the coating formed by 10 min in different
electrolyte solutions as a function of discharge current density.
current density, thus causing the rapid decrease of the specific ca-
pacitance. Moreover, the larger size of the hydrated cation causes a
higher ion diffusion resistance, and accordingly leads to a more
rapid decrease in the specific capacitance with increasing the dis-
charge current density. The ion diffusion resistance impact was
therefore emphasized by the high discharge current density that
makes the system far from equilibrium and highlights the kinetic
behaviors of the coatings, particularly thick ones. Further supportive
evidence for this hypothesis of ion diffusion resistance impact was
obtained when EIS of the same coating was recorded in different
electrolyte solutions. The small sine-wave amplitude ͑10 mV͒ and
broad frequency range ͑105–0.1 Hz͒ used for the EIS measurements
makes it possible to separate and study electrochemical processes
having different time constants, such as electrical conduction, ion
transfer, and capacitance. It can be seen that the EIS plots, shown in
Fig. 8, are similar in form, and can be divided into a high-frequency
component ͑inclined at 45°͒ and a low-frequency component ͑near
vertical͒. The intercepts of the plots with the real impedance axis
of the values of ⌬t with the different employed electrolytes, and
thereby reveal the dependence of the specific capacitance on the
metal ion nature of electrolyte, which is in accordance with the CV
measurements shown in Fig. 5. In order to explain this, one should
first understand the charge storage mechanism of manganese oxide
electrode used in neutral aqueous electrolyte solutions. Two main
mechanisms were proposed to explain the manganese oxide charge
storage. The first suggests the intercalation of protons or alkali metal
cations in the bulk of the material upon reduction, followed by de-
intercalation upon oxidation.18 The second is based on the surface
adsorption of alkali metal cations on manganese oxide.6 Recently, a
combination of the two explanations was made to get a better un-
derstanding of the charge storage mechanism in manganese oxide
electrodes.46 Most importantly, protons and alkali metal cations
from the electrolyte are confirmed to be involved simultaneously in
the charge storage process of manganese oxide thin-film electrodes,
and alkali metal cations are anticipated for charge compensation,
with an implication of alkali metal cations playing an important role
in the charge–discharge electrochemistry of manganese oxides.
Thermodynamically, the smaller alkali metal cation may get more
involved in charge compensation during the charge–discharge pro-
cess, causing high capacitance. This is in agreement with Lee and
Goodenough,6 who claimed that the larger potassium ion appears to
limit the capacity to chemisorption, and thus lowers the capacity to
about one-third that with a proton.
͑Z ͒ presents the uncompensated electrical resistance ͑R ͒ of the
Ј
⍀
coatings ͑Ra͒, the electrolyte ͑Re͒, and the electrical leads combined
͑Rl͒.47 The ͑almost͒ same intercepts of the plots suggest negligible
differences in electrical resistance of the electrolyte because the
Because the capability to discharge at a high rate is of great
concern for an electroactive material of supercapacitors, the charge–
discharge behavior of all the coatings grown in this work was also
examined at different current densities. Figure 7 presents the specific
capacitance of the coating formed by 10 min in the three above-
mentioned different electrolyte solutions as a function of the dis-
charge current density. The specific capacitances in all three electro-
lytes decrease gradually with increasing the discharge current
density. In contrast, the decrease in a 0.5 M LiCl solution is the
largest, but that in a 0.5 M KCl solution is the smallest. The specific
capacitance in a 0.5 M LiCl solution at a discharge current density
of 17.7 mA/cm2 drops by about 31.5% from that at 0.885 mA/cm2,
while the specific capacitance in a 0.5 M NaCl solution drops by
about 26.8%, and only 22.9% for that in a 0.5 M KCl solution. The
comparatively rapid deterioration of the specific capacitance in a 0.5
M LiCl solution may be attributed to the larger internal resistance of
the coating. One possible explanation is that the ion diffusion is not
able to reach all the available active sites in the coating, thereby
restricting the amount of material that can contribute to the mea-
sured capacitance. Dynamically, the ion diffusion resistance from
hydrated cations leads to a large ohmic drop at a high discharge
Figure 8. Impedance Nyquist plots of the coating formed by 10 min in
different electrolyte solutions.
Downloaded on 2014-05-11 to IP 205.208.51.61 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).